A serial data communication system includes a transmitter, a receiver, and a communication link or channel coupling the transmitter and receiver. The physical medium of the communication link is commonly either copper wire or optical fiber. In some types of communication systems, a data signal is sent on one channel, while a clock signal is sent on another channel in synchronism with the data signal. In other types of communication systems, only one channel is used. In this latter type of system, the transmitter sends the data signal on a channel without an accompanying clock signal. The receiver generates a clock signal from an approximate frequency reference and then detects the data stream by phase-aligning the generated clock signal to transitions in the data signal and sampling the data signal. Such a communication system is commonly referred to as a Clock and Data Recovery (CDR) system.
In CDR and similar communication systems, the quality of the received signal and thus the ability to recover the data at the receiver is often impaired by inter-symbol interference (ISI), crosstalk, echo, and other noise. In addition, impairments in the receiver itself may further degrade the quality of the received signal.
A technique known as peaking or pre-emphasis can be employed in a transmitter to enhance received signal quality and thereby enhance data recovery in an optical receiver. Peaking briefly boosts the output power of the transmitter and then returns the output power to a nominal level. For example, it is known that boosting or peaking transmitter power immediately following a signal level transition from a “0” bit to a “1” bit and, similarly, immediately following a transition representing a transition from a “1” bit to a “0” bit can enhance data recovery. In each instance, only the bit immediately following the transition is peaked.
In recent years, electrical data communication systems, in which an electrical signal transmitter and electrical signal receiver are coupled by a copper wire medium, have given way to optical data communication systems, in which an optical transmitter and optical receiver are coupled by an optical fiber. In optical data communication systems, the optical transmitter is commonly a laser, and the optical receiver is commonly a photodiode. The above-described peaking techniques have been carried over from electrical data communication systems to the newer optical data communication systems. Accordingly, it is known to boost laser power (e.g., by the laser driver circuitry adding a momentary peaking current) immediately following a bit transition.
The received signal can be represented by a graphical construct known as a “data eye,” “eye diagram,” or “eye pattern.” An eye diagram is a superposition of a number of impaired individual signals with varying frequency components (e.g., due to ISI and noise) over a unit interval (UI) of the data signal (i.e., the shortest time period over which the data signal can change state). As the various impairments increase, the quality of an eye diagram derived from or otherwise detected by observation of the received signal is impaired.
An eye diagram may be generated by applying the data signal to the vertical input of an oscilloscope or similar test instrument and triggering a horizontal sweep across the unit interval based on the data rate of the data signal. When the data signal corresponds to a pseudorandom data signal, the superposed samples appear on the oscilloscope display as an eye diagram with an eye opening bounded by two transition regions. Various features of the eye opening reveal information about the quality of the communication channel over which the serial data signal is transmitted. For example, a wide eye opening indicates that the data signal has a relatively low noise level and a relatively low bit-error rate, whereas a narrow eye opening indicates that the data signal has a relatively high noise level and a relatively high bit-error rate. It is known that the eye opening can be increased by employing the above-described peaking technique.
As illustrated in FIG. 1, it is known that a laser driven in a conventional manner (and without any added peaking) exhibits a natural or inherent peaking on the first bit after a bit transition, as a result of the laser's inherent self-relaxation oscillation. In FIG. 1, such natural or inherent peaking can be observed in a region 10 immediately following a bit transition from a “0” (low laser optical power) to a “1” (high laser optical power). That is, in region 10 the laser optical power briefly increases above a nominal power level 12 corresponding to a “1” before returning to the nominal power level 12 for the duration of the time the laser produces an output corresponding to a “1” bit. Similarly, natural or inherent peaking can be observed in a region 14 immediately following a bit transition from a “1” (high laser optical power) to a “0” (low laser optical power). That is, in region 14 the laser optical power briefly decreases below a nominal power level 16 corresponding to a “1” before returning to the nominal power level 16 for the duration of the time the laser produces an output corresponding to a “0”.
The desire for economical, high-throughput, low power, optical data communication systems has led to the use of multi-mode optical fiber and vertical cavity surface-emitting lasers (VCSELs). Considerations relating to semiconductor physics and manufacturing reliability impose a practical limit on VCSEL bandwidth. It would be desirable to facilitate increased data rates in such systems by enhancing received signal quality through means other than increased VCSEL bandwidth.